Practice variation in a longitudinal perspective: a multilevel analysis of the prescription of simvastatin in general practices between 2003 and 2009.

PURPOSE: The Dutch national guideline for Cardiovascular Risk Management advised in 2006 initiation of statin therapy in new patients with simvastatin. After the introduction of this guideline several measures were taken to increase the prescription of simvastatin in general practice. The aim of this paper is to describe the trend and interpractice variation in the relative share of simvastatin in statin prescription over a 7-year period (2003-2009) in Dutch general practices.
METHODS: Data were used from the National Registration Network of General Practice (LINH) in the Netherlands. We used multilevel logistic regression analysis with individuals nested within practices. We evaluated the variance at different time points by the intra-class correlation (ICC).
RESULTS: The share of simvastatin prescriptions of all statin prescriptions increased from 40% in 2003 to 82% in 2009. While the ICC was above 20% during the periods 2005 to 2007, it decreased rapidly from 25% in 2007 to only 9% in 2009.
CONCLUSION: Our results indicate that a combination of several intervention measures seems to have been successful. While it seems like the implementation of the guidelines alone had some effect, there was an additional effect when reimbursement measures were implemented. Furthermore, the practice seemed to have affected the adoption of the national CVRM guideline as the variation initially increased after the guideline implementation. By including information on both the proportion of simvastatin prescribed and interpractice variation, we achieved a more complete evaluation of how simvastatin prescription developed over time and how it was related to the different measures taken.